Improving Fantasy Football

[albionmoonlight]

So we all know that, based on the way fantasy football works, the value of players--while loosely connected to reality--differs from their actual value on the football field. If you need an example--note that picking A. Brooks or M. Bulger over T. Brady in a fantasy league before this season would not really have been considered a crazy play.

Fantasy football player rankings also differ from real life rankings based on things like having to start two running backs in your league and running backs being devalued if they have a goal-line specalist on their team to take away touchdowns, etc.

People, by and large, have learned the way that the fantasy game works and have adjusted their drafting accordingly. Now that there is a whole industry around fantasy scoring being the way it is--I don't see it changing.

But I wonder why that had to be. After the idea of fantasy football came into being and took off, why didn't we start to tweak it to make it more like real life? Why did we just conform to a system that we all know if flawed--after widespread use confirmed the flaws.

And if you could tweak fantasy scoring to make it more like reality--how would you do that?

[rkmsuf]

Because we already have real football.

Also because football statistics - as opposed baseball statastics -- are just a different animal. Winning and stats don't always go hand in hand in such a team game.

Without stats you can't have fantasty sports so in football you are forced to use what you have. Unless you come up with some real abstract valuations I'm not sure where you go with it to make it popular enough to be able to sit down on Sunday and follow the action.

[WSUCougar]

And if you could tweak fantasy scoring to make it more like reality--how would you do that?
This is probably in effect for some leagues, but I think there could be an "on winning team" factor or multiplier for each player. In other words, perhaps Brady's 2 TDs in a winning effort are worth more than Brooks 2 TDs losing.

[rkmsuf]

This is probably in effect for some leagues, but I think there could be an "on winning team" factor or multiplier for each player. In other words, perhaps Brady's 2 TDs in a winning effort are worth more than Brooks 2 TDs losing.

So how about this past Sunday? Tom Brady. Zero times something vs. Aaron Brooks and his 1 TD.

Or Maybe Peyton Manning vs. Kerry Collins.

[QuikSand]

Interesting topic.

[MikeVick7]

This is the deepest I've seen anyone ever dive into the realm of fantasy football.

[QuikSand]

One thought - traditional fantasy tends to value gross statistics. In theory, you might try to find some way to better value efficiency than that, to better get to this idea.

QB #1 -- 20/28 for 190 and 1TD, 0 Int
QB #2 -- 25/44 for 290 and 2TD, 3 Int

Which guy helped htis team more? Which guy gets more fantasy points?

[Passacaglia]

Don't most of you guys have scoring that penalizes incompletions? That's a start.

[QuikSand]

Don't most of you guys have scoring that penalizes incompletions? That's a start.

I would guess that fewer than 20% of fantasy leagues penalize for anything beyond turnovers.

(points or yards allowed by defenses and missed kicks notwithstanding)

[cuervo72]

Time of posession? Don't know if you can translate that to individual players or not...

[rkmsuf]

One thought - traditional fantasy tends to value gross statistics. In theory, you might try to find some way to better value efficiency than that, to better get to this idea.

QB #1 -- 20/28 for 190 and 1TD, 0 Int
QB #2 -- 25/44 for 290 and 2TD, 3 Int

Which guy helped htis team more? Which guy gets more fantasy points?

Buy isn't a major part of the wide appeal the simplicity of the game?

[albionmoonlight]

Buy isn't a major part of the wide appeal the simplicity of the game?
On some level, yes, but now there are guys who spend more time planning for their fantasy drafts then for the birth of their first child.

And there are people who really spend time considering who their #5 WR is going to be.

The things that those people look at (such as whether to pick the slot receiver on a team that likes to spread the field in the red zone, or whether to pick the #2 receiver on a team who might be breaking in a rookie QB by the end of the season) are not always what you would look at when really building a football team. I am asking if that can/should be changed and how.

Some ideas might include penalizing for incompletions and drops. Also decreasing for touchdowns vs. yardage.

What about bonuses for third down runs that result in a first down? Bonuses for third down passes and catches that result in a first down?

Some sort of multiplier to decrease the importance of running backs? The 8th best back in the league should not be being picked before the best quarterback. At least not in the world as I see it.

[QuikSand]

Buy isn't a major part of the wide appeal the simplicity of the game?

I'm not convinced that any of the ideas here would actually improve fantasy football in terms of interest or excitement -- I'm just following up on the idea of a higher correlation between quality fooball outcomes and quality fantasy football outcomes.

[Wolfpack]

I had thought on this some just recently (though I'm not partaking in fantasy football this year). Perhaps there should be some sort of bonus or multiplier for offensive players who score in close and late games? Say, 9 points instead of six for a touchdown when it's the fourth quarter and the margin is less than 10 points? That should put some emphasis on those who perform when things are on the line. Perhaps along those lines, consider a bonus point to anyone scoring TDs in a road game to indicate the tougher environment? Just some thoughts aloud.

[Samdari]

QB #1 -- 20/28 for 190 and 1TD, 0 Int
QB #2 -- 25/44 for 290 and 2TD, 3 Int

Which guy helped htis team more? Which guy gets more fantasy points?

The problem QS, is that while the answer to the above question seems obvious to you (or so I infer) people would debate even that. Some would argue the extra 100 yards and TD offset the 3 INT's. Others would argue the converse.

As for penalizing incompletions to more accurately measure a QB's value, how do you account for the plays that an incompletion is a good play? If the line fails to account for a guy, or completely misses a block, and is in the QBs face almost immediately, I'd rather my QB throw the ball away if he has time than do anything else. The best (most) QB's can do sometimes is throw an incompletion, or even take a sack.

Stats don't necesarily accurately measure the individual performance of a player. Witness Timmy Smith's 200+ yards in the Super Bowl. The line got him that - he was an unremarkable player. Yet his fantasy stats were killer, especially that day. Derrick Blaylock was a fantasy stud for a few games last season, despite being no more than a decent backup caliber player. The inherent flaw is in the stats, not really how those translate to points.

EDIT: Changed to a fancier word in a shameless attempt to impress QuikSand.

[cuervo72]

QB #1 -- 20/28 for 190 and 1TD, 0 Int
QB #2 -- 25/44 for 290 and 2TD, 3 Int

Which guy helped htis team more? Which guy gets more fantasy points?

Which gets player of the game honors?

:D

[QuikSand]

The problem QS, is that while the answer to the above question seems obvious to you (or so I infer) people would debate even that. Some would argue the extra 100 yards and TD offset the 3 INT's. Others would argue the converse.

Okay, fine - make the second guy 25/58 to make the effect even more obvious. Hoep you get the original point.

[QuikSand]

converse

*impressed*

[QuikSand]

The inherent flaw is in the stats, not really how those translate to points.

Okay, on this point.

Sure, there are problems with football stats reflecting what actually happened. Granted.

However, if you accept this as such a compelling flaw as to render any doscussion of stats as a proxy for perfomance, then you simply can't participate in this discussion. You are obliged to just give up, and declare that nothing can be done. Which is fine as a consistent philosophical position, but just not very interesting.

I'd prefer to work out the problems wherever we can (don't penalize the QB for dropped passes or undue pressure - okay, I'm on board), and then use what we have after we make our best efforts to retain its value.

[Samdari]

Okay, fine - make the second guy 25/58 to make the effect even more obvious. Hoep you get the original point.

You are still missing my point - no matter what you make the att/comp, some will still feel that the extra 100 yards and TD mean THAT guy helped his team more, no matter how convinced you are that it is obvious that QB1 was better.

[rkmsuf]

I'm on record in the "declare nothing can be done camp". At least nothing remotely interesting or logical.

[Subby]

Are there some stats that Football Outsiders uses to get a better handle on player value and performance? I would be interested in taking a more SABRmetric approach if that is even possible...

[QuikSand]

You are still missing my point - no matter what you make the att/comp, some will still feel that the extra 100 yards and TD mean THAT guy helped his team more, no matter how convinced you are that it is obvious that QB1 was better.

And I'd still disagree with them, I just made the argument more lopsided in my favor for that argument. Yup, I get your point. At some level, gross statistics clearly don't tell the whole story.

This becomes the same discussion that was held over the NL CY Young... on the value of incremental additional statistics. I would say that under any measure, a QB going 5 out of 30 for 100 yards, 2 TD and 3 interceptions is not a positive contribution to the team.

[rkmsuf]

Are there some stats that Football Outsiders uses to get a better handle on player value and performance? I would be interested in taking a more SABRmetric approach if that is even possible...


you keep that baseball nerd crap out of fantasy football

[Silver Owl]

One thought - traditional fantasy tends to value gross statistics. In theory, you might try to find some way to better value efficiency than that, to better get to this idea.

QB #1 -- 20/28 for 190 and 1TD, 0 Int
QB #2 -- 25/44 for 290 and 2TD, 3 Int

Which guy helped htis team more? Which guy gets more fantasy points?

Which player is on the winning team?

[VPI97]

QB #1 -- 20/28 for 190 and 1TD, 0 Int
QB #2 -- 25/44 for 290 and 2TD, 3 Int

Which guy helped his team more? Which guy gets more fantasy points?I'm not sure what you are looking for...there are already some FF systems that allow you to penalize incompletions and interceptions enough that QB#1 could be the higher scoring player. Hells bells, Kyle Orton scored me -11 points this week in our league.

[Samdari]

Okay, on this point.

Sure, there are problems with football stats reflecting what actually happened. Granted.

However, if you accept this as such a compelling flaw as to render any doscussion of stats as a proxy for perfomance, then you simply can't participate in this discussion. You are obliged to just give up, and declare that nothing can be done. Which is fine as a consistent philosophical position, but just not very interesting.

I'd prefer to work out the problems wherever we can (don't penalize the QB for dropped passes or undue pressure - okay, I'm on board), and then use what we have after we make our best efforts to retain its value.

I maintain that using official stats, that yes, no improvements can be made.

As for using the things you talk about to truly evaluate QB performance, I agree with using them (as well as giving credit for throwing the ball away when appropriate, or just taking the sack when the situation calls for it) to evaluate a QB's true performance. Personally, I would not credit a QB for yards after catch (but, again, others obviously disagree with me). Two problems here are that (1) there is no official stat for dropped passes, pressures, etc. and (2) such things are much more subjective than the things actual stats are counted for. All of the things they count as official stats are pretty objective. Sure, there is some subjectivity on whether a scrambling QB was sacked or ran for negative yards, but for the most part, you cross a yard line, you get credit for a yard gained. And, the stats are available real time, which is a great deal of the appeal to many people.

As a side note I believe that the current most commonly used system was arrived at not necesarily by trying to truly measure which players helped their teams more, but to balance scoring out among several different positions. If you count passing TDs/yds the same as rushing and recievings, QBs dominate the scores.

[Subby]

Incompletions are often the sign of a "good" decision by a qb - or an otherworldly play by the defense - or poor protection by an OL

[QuikSand]

So, maybe the ideal system would be closer to what college teams use for stickers on the helmet? Someone watches the game (film) and makes a subjective assessment, and decides that this guy gets three buckeyes, and that guy gets two, and that other guy gets none?

[Samdari]

And I'd still disagree with them, I just made the argument more lopsided in my favor for that argument. Yup, I get your point. At some level, gross statistics clearly don't tell the whole story.

You are STILL missing my point - different people value different things from football players. Its not that gross statistics don't tell the whole story, its that they tell a different story to different people.

There are a number of football fans that would shock you who do believe that yards and TDs are more important measures than completion and INT%.

to put it to ridiculous extremes, yes, some people would value

5/30 for 200 yds, 2 TD, 3 INT

over 12/19 100 yds, 1TD, 0 INT

Since the first guy moved his team more and scored more points for his team. And with some justification - TDs are more highly correlated to winning than turnovers.

I am not saying I agree with that, but I know there are many people who do. What I am trying to say is that when you get into more subjective or esoteric (vocabulary bonus points) ways to measure performance, you'll never get agreement on how to score such a league (what should be count + or -) let alone find a way to consistently collect/measure such data. Any system you came up with, you'd still have people arguing, whether it made Brady, Manning, or Vick the #1 QB, you would have (fantasy) players who value the other guy arguing how flawed the system is, since there is such a wide range of opinions on who the best QB is. That means there is a pretty wide range of opinions on what is important. That's your trouble here.

[QuikSand]

to put it to ridiculous extremes, yes, some people would value

5/30 for 200 yds, 2 TD, 3 INT

over 12/19 100 yds, 1TD, 0 INT

Since the first guy moved his team more and scored more points for his team. And with some justification - TDs are more highly correlated to winning than turnovers.

The world has irrational people. I am quite comfortable making judgments about right and wrong or better and worse, even recognizing that some such people will disagree with my conclusions, even under extreme circumstances. So it goes.


Since this conversation (neither started nor originally framed by me) is about trying to make fantasy outcomes more in line with real football outcomes... I'm doing my best with the essential fuel for fantasy fooball, the statistical outcomes from games.

I realize it's imperfect to do so. I also would suggest that it's not so vastly imperfect as to render the entire exercise completely pointless.



On one specific point -- yes, scoring touchdowns correlates highly to winning games. But I'm not sure that the method of such touchdowns is all that important. To use your "silly example of misleading stats" theory in reverse -- which QB helped his team win more, the guy who completed 7 of 7 passes to get the team down to the 1 yard line and then handed it off for an easy TD plunge, or the guy who tossed the 1-yard TD pass after watching the running game get them all the way down the field?

[Samdari]

The world has irrational people. I am quite comfortable making judgments about right and wrong or better and worse, even recognizing that some such people will disagree with my conclusions, even under extreme circumstances. So it goes.


Since this conversation (neither started nor originally framed by me) is about trying to make fantasy outcomes more in line with real football outcomes... I'm doing my best with the essential fuel for fantasy fooball, the statistical outcomes from games.

I realize it's imperfect to do so. I also would suggest that it's not so vastly imperfect as to render the entire exercise completely pointless.



On one specific point -- yes, scoring touchdowns correlates highly to winning games. But I'm not sure that the method of such touchdowns is all that important. To use your "silly example of misleading stats" theory in reverse -- which QB helped his team win more, the guy who completed 7 of 7 passes to get the team down to the 1 yard line and then handed it off for an easy TD plunge, or the guy who tossed the 1-yard TD pass after watching the running game get them all the way down the field?

I have no doubt that you could come up with a system that would satisfy you as to correlating between "real" performance and awarding appropriate fantasy points. But, since fantasy football is boring to play alone, you would also need to convince others that what you find important is what is really important. THAT, my friend, is the futile part of this exercise.

As for the details in your last paragraph, you are focusing too much on them and not on the meat of my argument - that you will never convince others of how right you are. BEING right is not enough here. I actually agree with your view of the example you give. You don't need to convince me, you just need to convince everyone who plays fantasy football. Good luck and keep us updated on your progress with that.

[QuikSand]

Here's the question, as framed by the topic-starter:

And if you could tweak fantasy scoring to make it more like reality--how would you do that?

Apparently, Samdari, you now suggest that the appropriate goal is not, in fact, to make fantasy scoring more like reality... but rather to make it more like an interpretation of reality that each and every football fan would agree with. That seems, to me, a higher standard than was initially suggested, or than anyone (myself included) ever aspired to.

So, if that is your point -- and that is the only acceptable success in your mind, to find something that everyone would agree with -- then yes, I did indeed miss your point, and you are also correct that my various suggestions and thoughts in this thread fall short of that particular goal.

I guess that means you win.

[John Galt]

I guess that means you win.

I disagree. But, then again, I use a different scoring system.

[albionmoonlight]

FWIW, what I see as the ideal of this discussion is trying to get the barstool debate: "Who are the top ten fantasy players in the league right now" to mirror as closely as possible the barstool debate "If you were starting an NFL team for this season, who are the top ten guys you would pick?"

We can all agree that there are guys who, if they were picking the QB to lead their team, would argue for Manning, or they would argue for Culpepper, or they would argue for Brady, or they would argue for Vick.

We don't all agree on what makes the best football players. We don't all agree on what makes the best fantasy football players. But we use a different set of variables when making those distinctions.

Certainly a guy who thinks that Vick is a better football player than Brady will come up with a different fantasy scoring system in response to my question than someone who thinks the converse. But we can all agree (I hope) that the system which we have now does not really reflect anyone's reality. Is Tiki Barber really a worse player this year just because the Giants replaced Ron Dayne with a real goal line back?

[bryce]

efsports.com

rulebook = http://dynasty.efsports.com/rules/EFS_Rules_2004.pdf

Though I have never played their league, at first glance they have easily the most comprehensive fantasy football system I've ever seen that attempts to mirror the NFL as closely as possible.

edited for link

[Raiders Army]

So we all know that, based on the way fantasy football works, the value of players--while loosely connected to reality--differs from their actual value on the football field. If you need an example--note that picking A. Brooks or M. Bulger over T. Brady in a fantasy league before this season would not really have been considered a crazy play.

Fantasy football player rankings also differ from real life rankings based on things like having to start two running backs in your league and running backs being devalued if they have a goal-line specalist on their team to take away touchdowns, etc.

People, by and large, have learned the way that the fantasy game works and have adjusted their drafting accordingly. Now that there is a whole industry around fantasy scoring being the way it is--I don't see it changing.

But I wonder why that had to be. After the idea of fantasy football came into being and took off, why didn't we start to tweak it to make it more like real life? Why did we just conform to a system that we all know if flawed--after widespread use confirmed the flaws.

And if you could tweak fantasy scoring to make it more like reality--how would you do that?
I think you want pickem...with or without the point spread.

[Gallifrey]

I have run a league for 12 years now, and the rule I like best in my league is your D/ST will lose a point for every one given up.
So last night the Giants D/ST finished at a -42. As in the NFL, you need to make up the points your Defense gives up.

This has created great games over the years. You get 2 points for a sack. 3 for a TO. 6 for a D/ST TD.
I lost a championship game on a 102 yard kickoff returned for a TD against MY D/ST. I went from up by 12 to loosing. Then on the kickoff with less than 20 seconds left in the game my return man was dropped by the kicker...the last guy! I was on a table yelling. Great fun.

[TRO]

I would say that one reason fantasy football is so popular is that the scoring is not overly complex. Anything that is easy to pick up will make it inherantly more popluar. Sure, you have people that spend months planning their draft, but those same people are generally in customizable leagues or in pickup leagues to take advantage of the casual fan.

Really, does a diehard fantasy player really want to complicate the world of fantasy football and potentially scare way, to borrow a poker term, the fish?

[TRO]

I have run a league for 12 years now, and the rule I like best in my league is your D/ST will lose a point for every one given up.
So last night the Giants D/ST finished at a -42. As in the NFL, you need to make up the points your Defense gives up.

This has created great games over the years. You get 2 points for a sack. 3 for a TO. 6 for a D/ST TD.
I lost a championship game on a 102 yard kickoff returned for a TD against MY D/ST. I went from up by 12 to loosing. Then on the kickoff with less than 20 seconds left in the game my return man was dropped by the kicker...the last guy! I was on a table yelling. Great fun.
So you force a team to not have a roster vacancy at DEF then?

[jamesUMD]

this is a stretch, but the technology is almost there. If you have a starter go out, your backup immediately comes in and accrues stats from the time that the starter went out. Drives me nuts when my QB gets hurt in the beginning of the game and I have to go the rest of the way w/o a QB.

Being able to have back-ups. Not reserve players mind you, but guys where you can get points from. Maybe something where you set the starters playing time like on a scale of 1 to 4, with 4 giving your starter 100% of their points, and a 2 gave 50% of the starters points and 50% to your back-up at the position. That way you might be able to parlay a deeper team vs. the team that has 3 or 4 absolute studs.

My league up until this year used a back-up RB, WR, and TE that got half points (scoring only league) of the starters which was cool because it was not as hit or miss. For example this week, of this year, I started Deuce McAlister and Steven Jackson, and had Brian Westbrook on the bench. I got nothing from Westbrook because of it. In past years he would have been my 3rd RB and I at least would have gotten 1/2 his point total. It makes team depth more important.

We just can't find a work around for our rules we made with CBS Sportsline. Just my 2 cents.

[KWhit]

FWIW, what I see as the ideal of this discussion is trying to get the barstool debate: "Who are the top ten fantasy players in the league right now" to mirror as closely as possible the barstool debate "If you were starting an NFL team for this season, who are the top ten guys you would pick?"
Exactly. I think the first thing to be done is to increase the importance of the QB in fantasy football. It was mentioned above, but I think it bears repeating that the QB is the most important single player on the field and should be valued as such in fantasy football.

[sterlingice]

This is why, in any league I start, there are 2 QBs. Right now, most fantasy football leagues are content to make it a scramble for RBs. At least with 2 QBs, you have QB as important, if not moreso, than RB. I've always thought it incredibly stupid that the 12th best QB in a 10 team league is worth roughly one of the bottom 5 RBs in the league.

SI

[ISiddiqui]

Exactly. I think the first thing to be done is to increase the importance of the QB in fantasy football. It was mentioned above, but I think it bears repeating that the QB is the most important single player on the field and should be valued as such in fantasy football.
In the leagues I'm in, completions count as one point. In one league incompletions are -0.5 point while in the other, incompletions aren't a negative. They have the effect of making the QB THE most important player.

Also making only 1 RB as mandatory would help as well. The reason RBs are valued so highly is because of scarcity.

I have to agree with jamesUMD though that there should be some backup capability mid fantasy game. Whether it is 1/2 points or substitutions for game-ending injury.

[sterlingice]

There are two problems, however, and one is diametrcially opposed to what you're looking at (#2).

1) If we went by the thread about how to build a football team, then pancakes would be the most important stat followed closely by hurries and sacks. And, really, no one follows offensive and defensive line like they do skill position players with countable stats.

2) If you made having the 10th best QB in the league more important than having, say, the best RB or best WR or something close to that, it would be like it is now- RBs are the only thing with any value (except now it would be QBs). Basically, if you were 1st in the first round, you'd have a huge advantage, barring injury. Who cares if you have the best RB, you're screwed because you have the 10th best QB.

SI

[INDalltheway]

I think yardage should be increased and TDs decreased... Maybe QBs should get 1 point for a drive that results in FG? I just hate how 18 yards and 1 TD beats a guy with 104 yards and 0 TDs...

[Cringer]

efsports.com

rulebook = http://dynasty.efsports.com/rules/EFS_Rules_2004.pdf

Though I have never played their league, at first glance they have easily the most comprehensive fantasy football system I've ever seen that attempts to mirror the NFL as closely as possible.

edited for link

thats some detailed stuff, I admit i didn't come close to looking through everything.

One I have always liked is Scoresheet Sports. Their football is all I have ever done. I started playing with them back in high school for a few years, then did it for a 2 or 3 seasons in a row a few years ago. I like them because you get individual defensive players, because taking a whole defense has always bugged me.

Here is their scoring system...

http://www.scoresheet.com/football/fb_cover.html

just scroll down a little

[Daimyo]

If you want to make fantasy football better confirm to real life you need to figure out approximate values in terms of pts that each event in a game is worth and then assign them to fantasy players accordingly. Its not as easy as baseball because so many players are involved to make each "event" happen, but I imagine with the all historical data it would be possible. The Hidden Game of Football would be a great start and footballoutsiders.com has kind of taken up that flag more recently...

[henry296]

Perhaps moving from a points system to a roto system that is used in Head to Head baseball and hockey leagues.

You could have 10 categories such as:

Passing Yards
Rushing Yards
Receiving Yards
TDs
Completion%
Yards per Rush
Yards per Reception
Yards per Attempt
Net Turnovers (Defensive TOs - Offensive TOs)
Points Allowed
Sacks